Tidy Wine Data

Import winemag-data_first150k.csv data.

wine150_df = 
  read_csv("./data/winemag-data_first150k.csv") %>% 
  janitor::clean_names() %>% 
    select(id = x1, points, price, country, province, variety, winery) %>% 
  mutate(
        country = factor(country),
        province = factor(province),
        variety = factor(variety),
        winery = factor(winery)     
    )

## Warning: Missing column names filled in: 'X1' [1]

The original dataset winemag-data_first150k.csv has 150930 observations.
The numbers of missing values for some important variables are as follows.
id: 0 missing observations.
points: 0 missing observations.
price: 13695 missing observations.
country: 5 missing observations.
province: 5 missing observations.
variety: 0 missing observations.
winery: 0 missing observations.

Create average points and price at the country and province levels respectively for winemag-data_first150k.csv data.

wine150_tidy =
  wine150_df %>% 
  arrange(country, points) %>% 
  group_by(country) %>% 
  mutate(
    points_avg_country = round(mean(points), 2),
    points_med_country = median(points),
    price_avg_country = round(mean(price, na.rm = TRUE), 2),
    price_med_country = median(price)
  ) %>% 
  ungroup(country) %>% 
  arrange(province, points) %>% 
  group_by(province) %>% 
  mutate(
    points_avg_province = round(mean(points), 2),
    points_med_province = median(points),
    price_avg_province = round(mean(price, na.rm = TRUE), 2),
    price_med_province = median(price)
  ) %>% 
  ungroup(province) %>%  
  arrange(country, province, points)

Due to 5 missing values of country and province in winemag-data_first150k.csv data, average points and price are also created for this missing group. Caution is needed for subsequent analysis.

Create average points and price at the variety and winery levels respectively for winemag-data_first150k.csv data.

wine150_tidy =
  wine150_tidy %>% 
  arrange(variety, points) %>% 
  group_by(variety) %>% 
  mutate(
    points_avg_variety = round(mean(points), 2),
    points_med_variety = median(points),
    price_avg_variety = round(mean(price, na.rm = TRUE), 2),
    price_med_variety = median(price)
  ) %>% 
  ungroup(variety) %>% 
  arrange(winery, points) %>% 
  group_by(winery) %>% 
  mutate(
    points_avg_winery = round(mean(points), 2),
    points_med_winery = median(points),  
    price_avg_winery = round(mean(price, na.rm = TRUE), 2),
    price_med_winery = median(price)
  ) %>% 
  ungroup(winery) %>%    
  arrange(variety, winery, points)

save(wine150_tidy, file = "./data/wine150_tidy")

There is not any missing value of variety in winemag-data_first150k.csv data.